Clustering of time-course gene expression data using a Bayesian infinite mixture model based approach